Tutorials

Raziq Yaqub

Niksun Inc. Princeton, New Jersey, USA

 

Title Cyber Security for Critical Physical Infrastructure in Pakistan
Abstract

Media, research, and statistics reveals that cyber warfare is currently being weighed actively in the region where Pakistan is located. Dawn, May 19, 2015 (http://www.dawn.com/news/1182856) notifies that “US network security firm FireEye tracked attacks on government, defense and educational infrastructure in at least 26 different countries, including Pakistan”. Eugene Kaspersky, CEO of Kaspersky Labs said, “Critical infrastructures – power, healthcare, communications – are all under threat” (http://www.scmagazineuk.com/eugene-kaspersky-interview-critical-infrastructure-is-under-threat/article/399530/).In last couple of years, Pakistan experienced several types of Cyber Security attacks and threats on power, telecommunications, and several other industrial automation systems. Thus it is imperative that Pakistan takes institutional steps to protect Critical Physical Infrastructure. Government of Pakistan has already taken some substantial steps, e.g., formation of focal ministry to handle exclusively the Cyber Security issues. Creation of Cyber Security Task Force with the affiliation of Ministry of Defense, Ministry of IT, Ministry of Foreign Affairs, and Ministry of Information. These steps were taken as a result of the Cyber Security Council Bill that was presented by Senator Mushahid Hussain Sayed with the aim of creating policy, guidelines and strategy on cyber security issues according to international best practices. Though, Cyber Security have been recognized as an emerging challenge by the Government of Pakistan, and necessary steps have already been taken, the speaker believes that strategy for Pakistan’s Cyber Security must also include “Cyber Security for Critical Physical Infrastructures (CPI)”, because it is extremely critical. The argument being; that a stolen trade secret may lead to IPR leakage to lost sales, and to lost jobs, however, a compromised CPI may have devastating impacts on humanity, and environment. For example launching an attack on the Critical Physical Infrastructures, may cause flooding the dams causing massive deaths, and high level damage, destruction or disruption that a traditional military attack would cause. This tutorial would begin with the definition of Cyber Security for CPI. CPI includes Power Grid, Nuclear Plants, Oil and Gas Refineries, Water Supply System, Emergency Services Infrastructure, Transport, Telecommunication infrastructure, Manufacturing and other Industrial Control Systems (ICS). The tutorial would also emphasize on how the tight coupling between information and communication technologies (ICT) and CPI introduces new security concerns, and provides adversaries with powerful tools and makes it vulnerable to new security challenges. The agenda of the tutorial would also include (a) Real insight into Cyber Security Concepts, (b) Security Requirements for ICS; in contrast to ICT, (c) Architecture of Industrial Control Systems (ICS) including Distributed Control Systems (DCS) and SCADA, (d) Emerging Threats and Vulnerabilities in CPI Specific to Pakistan because of its Location and Strategic Role, (e) Failure of Interlocks, Permissive and Remediation, and (f) Recommendations to Address the Challenges, specially the challenge areas where previous recommendations never existed or not been fully addressed. The tutorial will also provide information on what tools, capabilities, and services are available, and what are needed by Pakistan. The talk will also emphasize the need of public awareness programs that could convey the doctrine of Cyber Security Council of Pakistan. The author believes that common public does not understand impact of a massive Cyber-attack on Critical Physical Infrastructure; that may result the potential for large-scale power outages or man-made environmental disasters and cause physical damage, loss of life and other cascading effects, which may far worse impacts compared to stealing confidential information. The talk will also provide food for thought for university students and researchers on how to explore innovative techniques and conduct Cyber Security related research specific to different silos, such as Government, Defense, Financial Institutions, ICT Infrastructure, and Industrial Control System (including Power Industry’s SCADA system). Thus, the talk will be beneficial for the government policy makers, industry leaders and decision makers, corporate and academic researchers, professors and students, executives from telecom agencies such as Pakistan Telecom Authorities (PTA), Utility Industry such as WAPDA, and Alternative Energy Development Board.

Bio

Dr. Yaqub earned a Ph.D. in Wireless Communication from Keio University, Japan, and MBA in Marketing from Fairleigh Dickenson University, USA. He is one of the pioneers of LTE/4G, and an inventor of numerous technologies of 4th Generation Wireless Communication, and Smart Grid. He received “Inventor of the Year Award” from the Governor of the State of New Jersey, USA, through Inventors Hall of Fame. Dr. Yaqub remained an Executive Director of Toshiba America Research, Sr. Consultant to the State of New Jersey, a spokesperson in 3GPP on behalf of Department of Homeland Security”, an Associate Professor at University of Tennessee, and a Director of Technical Training, for NIKSUN, a global leader in Cybersecurity, headquartered in Princeton, New Jersey. Dr. Yaqub is also skilled in teaching, conducting research, developing new graduate level courses, and building industry-academia collaboration. Some of the new courses he developed and delivered include “Smart Grid”, “4G Networks”, and “Advanced Metering Infrastructure and Cyber Security”. He has been teaching graduate students of CECOS University Peshawar online via Skype. He is an approved Ph.D. Supervisor of HEC Pakistan, and is invited to Pakistan as a Resource Person or a Visiting Professor by HEC Pakistan, or Pakistan National Talent pool. His research interest is in 4G, Smart Grid technologies, (including Electric Vehicles, Wind Generation, Smart Home Energy Management, etc.), and Cyber Security. He filed 34 patents (18 already issued), published numerous papers in international conferences, and submitted 150+ contributions in technical standards organizations. He remained a working group chairman in Mobile Wireless Internet Forum, Chairman IEEE Membership Development, Rapporteur in 3GPP, keynote speaker, panelist, and guest speaker in numerous International conferences.

Safdar Ali and Abdul Majid

Pakistan Institute of Engineering and Applied Sciences

Safdar Ali                 Abdul Majid
Title Recent Development in the Breast Cancer Prediction models using Computational Intelligence Approaches
Abstract

Cancer is one of the rapidly growing diseases in the world. Cancer is the second largest cause of death after heart-related diseases. Breast cancer is the second major cause of worldwide cancer-related deaths. Both men and women are afflicted by breast cancer though it is common in women. Similar to other cancers, identification of cancerous tissues in early stages provides higher chances of survival and is helpful in avoiding unnecessary toxicity. Consequently, development of an efficient and reliable cancer prediction system is highly desirable. Such computational intelligent based prediction system has the capability to provide adequate information for cancer diagnosis and drug discovery. A clinician or academia or practitioners or technicians could easily utilize such prediction system for the diagnosis of cancer using protein sequences of the affected tissue. In the recent literature, researchers have proposed several intelligent approaches for the early diagnosis, prognosis and prediction of human breast cancer. These approaches used molecular signature based feature spaces for the prediction of breast cancer. Molecular signatures are predicted using basic knowledge from biosciences related fields of proteomics, genomics, etc. Proteins of a tissue would reflect the initial changes caused by the successive genetic mutations, which lead to cancer. Such changes/mutations would be exploited for the early diagnosis of breast cancer. These mutations of amino acid in protein sequences might be useful for drug discovery and cancer prediction. Descriptors are generated by statistical and/or mathematical methods using physicochemical properties of amino acids of protein related to cancer. Such descriptors are very beneficial for the diagnosis of breast cancer and could be used for the development of cancer prediction system. Various computational intelligence based individual and ensemble prediction models are developed for the detection and prediction of breast cancer. The cancer prediction problem can be approached by using both single and hybrid machine learning algorithms. In recent study, it is proved that SVM offered better performance over boosting, decision trees, KNN, neural networks, and random forest for cancer prediction. It is still a challenging task to develop a high performing ensemble system for cancer prediction. Earlier, Boosting, Bagging, and Random Forest based ensemble systems are developed by generating a set of classifiers that are trained from single learning algorithm. These approaches attempt to improve the prediction performance by iteratively retraining the base classifiers with a subset of most informative training data. These approaches have limited performance due to small number of biological samples and class imbalance. Another limitation is that these approaches have merely one level by taking the original input data to give single output prediction. On the other hand, recent classifier stacking based ensemble system generated by a set of classifiers that are trained from different learning algorithms has shown improved performance.

Bio

Dr. Safdar Ali

Dr. Safdar Ali received his M.Sc. degree in Physics from University of the Punjab, Pakistan and his Ph.D. degree in computer and information science from the Pakistan Institute of Engineering and Applied Sciences (PIEAS). His current research interests include artificial intelligence, machine learning, pattern recognition, image processing, and development of intelligent decision support system, as well as applications in human breast and colon cancers. In particular, Dr. Ali is interested in the development of computational approaches based ensemble systems for cancer and biomedical images.

Dr. Abdul Majid

Dr. Abdul Majid, Associate Professor, has a distinguished 23 years career in research, development, and teaching in Department of Computer and Information Sciences at PIEAS, Islamabad. He received MSc degree in Electronics from Quaid-i-Azam University in 1991. He obtained his MS and PhD degrees in Computer Systems Engineering from GIK Institute, in 2003 and 2006, respectively. He has completed his Post-Doc Research in the Department of Mechatronics, GIST, South Korea, in 2010. His research areas include pattern recognition, image processing, machine learning, and computational material science. His prominent research includes the development of intelligent decision support system for human breast and colon cancers. He is on the panel of reviewers of internationally reputed journals in the field of computer science. Under his supervision/co-supervision, 4 PhD, 14 MS and 12 BS students have completed their research thesis. Currently, three PhD students are carrying out research under his supervision. He is serving as a member of board of studies, board of faculty, and synopsis review committee. PAS, HEC, and PCST have acknowledged his research by awarding PAS Research Awards in Computer Science & IT in 2014, Outstanding Research Awards in 2013, and Research Productivity Awards in years 2011 and 2012, respectively.